PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Fast Computation of Graph Kernels
S V N Vishwanathan, Karsten Borgwardt and Nicol Schraudolph
In: Advances in Neural Information Processing Systems (2007) MIT Press , Cambridge, MA , pp. 1449-1456.

Abstract

Using extensions of linear algebra concepts to Reproducing Kernel Hilbert Spaces (RKHS), we define a unifying framework for random walk kernels on graphs. Reduction to a Sylvester equation allows us to compute many of these kernels in O(n^3) worst-case time. This includes kernels whose previous worst-case time complexity was O(n^6), such as the geometric kernels of Gartner et al and the marginal graph kernels of Kashima et al. Our algebra in RKHS allow us to exploit sparsity in directed and undirected graphs more effectively than previous methods, yielding sub-cubic computational complexity when combined with conjugate gradient solvers or fixed-point iterations. Experiments on graphs from bioinformatics and other application domains show that our algorithms are often more than 1000 times faster than existing approaches.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Book Section
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:3990
Deposited By:S V N Vishwanathan
Deposited On:25 February 2008